Vaccitech (United Kingdom)

TPS5119 Background: VTP-850 is a heterologous prime-boost immunotherapeutic consisting of 2 nonreplicating viral-vectored components: ChAdOx1-PCAQ, based on an adenoviral vector, and MVA-PCAQ, based on a modified vaccinia virus Ankara (MVA) vector. Both components encode the same 4 prostate cancer antigens: prostate-specific antigen (PSA), prostatic acid phosphatase (PAP), six transmembrane epithelial antigen of prostate 1 (STEAP1), and 5T4, an oncofetal antigen. Preclinical studies in inbred, outbred, and HLA-A2 transgenic mice show that VTP-850 is highly immunogenic. Immune responses were measured in splenocytes using IFN-γ ELISpot assay, multiparameter flow cytometry, and targeted in vivo killing assays. The results indicate that VTP-850 can elicit T cell responses to each of the 4 encoded antigens, with the responses to PSA of greatest magnitude in all mouse strains studied. Intravenous administration of the boost resulted in a 6-fold increase in the magnitude of antigen-specific T cells induced and increased in vivo killing relative to the intramuscular boosting route1. Methods: This is a first-in-human multicenter Phase 1/2 trial to evaluate safety, PSA response rate and duration, and immunogenicity of VTP-850 in men with biochemical recurrence of prostate cancer after definitive local therapy. Phase 1 (15-18 participants) will follow a 3+3 design to determine the recommended phase 2 regimen; dose level of both ChAdOx1-PCAQ and MVA-PCAQ, and route of administration of MVA-PCAQ (IM or IV). Phase 2 will consist of 2 stages. In Stage 1, 19 additional participants will be enrolled at the chosen Phase 2 regimen. If 4 or more of the 25 participants (including the 6 Phase 1 participants who received the same regimen) have a PSA response (defined as ≥50% reduction in serum PSA compared to baseline at any time, measured twice consecutively, at least 2 weeks apart), Stage 2 will enroll 100 additional participants. Participants will be followed for 6 months or until start of new therapy (e.g. ADT) or until development of metastatic disease. Participants who have a PSA response during the 6 months follow up will be followed for up to an additional 18 months. Patients who have undergone primary therapy for prostate cancer and have biochemical recurrence are eligible. Patients must have nonmetastatic (M0) disease; serum PSA of > 0.3 ng/mL for participants with prior radical prostatectomy or serum PSA of 2 ng/mL above nadir for participants with prior external beam radiation or brachytherapy; PSA doubling time ≤12 months; and testosterone > 75 ng/dL. They cannot have received ADT within 6 months prior to Day 1 and cannot have received prior chemotherapy, immunotherapy or experimental agent for prostate cancer. The trial is open in the USA, with sites in Spain and Italy planned to open later in 2023. 1. Vardeu A et al 2022. Clinical trial information: NCT05617040 .

Mathematical modelling methods and adaptive trial design are likely to be effective for optimising vaccine dose but are not yet commonly used. This may be due to uncertainty with regard to the correct choice of parametric model for dose-efficacy or dose-toxicity. Non-parametric models have previously been suggested to be potentially useful in this situation. We propose a novel approach for locating optimal vaccine dose based on the non-parametric Continuous Correlated Beta Process model and adaptive trial design. We call this the ‘Correlated Beta’ or ‘CoBe’ dose optimisation approach. We evaluated the CoBe dose optimisation approach compared to other vaccine dose optimisation approaches using a simulation study. Despite using simpler assumptions than other modelling-based methods, we found that the CoBe dose optimisation approach was able to effectively locate the maximum efficacy dose for both single and prime/boost administration vaccines. The CoBe dose optimisation approach was also effective in finding a dose that maximises vaccine efficacy and minimises vaccine-related toxicity. Further, we found that these modelling methods can benefit from the inclusion of expert knowledge, which has been difficult for previous parametric modelling methods. This work further shows that using mathematical modelling and adaptive trial design is likely to be beneficial to locating optimal vaccine dose, ensuring maximum vaccine benefit and disease burden reduction, ultimately saving lives

Optimal vaccine dosing is important to ensure the greatest protection and safety. Analysis of dose-response data, from previous studies, may inform future studies to determine the optimal dose. Implementing more quantitative modelling approaches in vaccine dose finding have been recently suggested to accelerate vaccine development. Adenoviral vectored vaccines are in advanced stage of development for a variety of prophylactic and therapeutic indications, however dose-response has not yet been systematically determined. To further inform adenoviral vectored vaccines dose identification, historical dose-response data should be systematically reviewed. A systematic literature review was conducted to collate and describe the available dose-response studies for adenovirus vectored vaccines. Of 2787 papers identified by Medline search strategy, 35 were found to conform to pre-defined criteria. The majority of studies were in mice or humans and studied adenovirus serotype 5. Dose-response data were available for 12 different immunological responses. The majority of papers evaluated three dose levels, only two evaluated more than five dose levels. The most common dosing range was 107–1010 viral particles in mouse studies and 108–1011 viral particles in human studies. Data were available on adenovirus vaccine dose-response, primarily on adenovirus serotype 5 backbones and in mice and humans. These data could be used for quantitative adenoviral vectored vaccine dose optimisation analysis.

Initial release of fully reproducible repository, with model weights available.